ARTICLE IN PRESS
1352-2310/$ - se
doi:10.1016/j.at
�Correspond
fax: +34934110
E-mail addr
Atmospheric Environment 39 (2005) 6409–6419
www.elsevier.com/locate/atmosenv
Size fractionation in mercury-bearing airborne particles(HgPM10) at Almaden, Spain: Implications for inhalation
hazards around old mines
Teresa Morenoa,�, Pablo Higuerasb, Tim Jonesc, Iain McDonaldc, Wes Gibbonsd
aInstitute of Earth Sciences ‘‘Jaume Almera’’, CSIC, C/Lluis Sole i Sabarıs s/n, Barcelona 08028, SpainbDepartamento de Ingenierıa Geologica y Minera, Universidad de Castilla-La Mancha, Plaza Manuel Meca 1,
13400 Almaden (Ciudad Real), SpaincSchool of Earth, Ocean and Planetary Sciences, Cardiff University, Cardiff CF10 3YE, Wales, UK
dAP 23075, Barcelona 08080, Spain
Received 8 April 2005; received in revised form 4 July 2005; accepted 18 July 2005
Abstract
Almaden has a 42000y mining history and an unprecedented legacy of mercury contamination. Resuspended
airborne particles were extracted from mine waste (Las Cuevas), retort site soil (Almadenejos), and urban car park dust
(Almaden), separated into fine (PM10) and coarse (PM410mm) fractions, analysed for mercury using ICP-MS, and
individual HgPM characterised using SEM. Cold extractable mercury concentrations in PM10 range from 100 to
150 mg g�1 (car parks), to nearly 6000 mg g�1 (mine waste), reaching a world record of 95,000 mg g�1 above the
abandoned retort at Almadenejos where ultrafine HgPM have pervaded the brickwork and soil and entered the food
chain: edible wild asparagus stem material from here contains 35–65 mg g�1 Hg, and pig hair from animals living,
inhaling and ingesting HgPM10 at the site yielded 8–10 mg g�1. The PM10 fraction (dusts easily wind transported and
deeply inhaled) contains much more mercury than the coarser fraction. The contribution of HgPM10 to ecosystem
contamination and potential human health effects around old mercury mines has been underestimated.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Almaden; Mercury aerosol contamination; PM10
1. Introduction
The Almaden area, 285 km south of Madrid in south-
central Spain (Fig. 1), has supplied more than one-third
of all mercury ever mined and refined by humans
(Saupe, 1990; Ferrara et al., 1998; Higueras et al., 2003).
This is the largest point source of mercury in the world,
e front matter r 2005 Elsevier Ltd. All rights reserve
mosenv.2005.07.024
ing author. Tel.: +34934095410;
012.
ess: [email protected] (T. Moreno).
as illustrated by the fact that 90% of all dental amalgam
used in the European community was mined here
(Wilman, 2004). Primary ore has been extracted from
both underground and open-cast mines, and processed
locally to produce a huge quantity of the liquid metal: an
estimated 8 million flasks (250,000 tonnes). Mining at
Almaden has now ceased, but the unprecedentedly long
history of mining and processing has left a spectacular
legacy of contamination that poses a health threat to the
current population and to the local ecosystem (Higueras
et al., 2005). Studies have recently been published on
d.
ARTICLE IN PRESS
Fig. 1. Map of Almaden area showing geology and geographic location of population centres, main mines, and the Almadenejos
retorting site with rose diagram showing main wind directions in the area. Samples were collected from the mining sites (shown as
symbol on map) at Las Cuevas, El Entredicho and the Retorting Plant at Almadenejos, and from two sites within the Almaden city
centre (University and Supermarket car parks).
T. Moreno et al. / Atmospheric Environment 39 (2005) 6409–64196410
atmospheric Hg concentrations (20,000 ngHg m�3 in the
mining complex in Almaden town, 213.7 ngHg m�3
mean value for Almaden district; Ferrara et al., 1998;
Higueras et al., 2005), amounts of aquatic mercury in
local streams and rivers (Berzas Nevado et al., 2003;
Gray et al., 2004), and residues in soils and plants near
mines and processing sites (Higueras et al., 2003, 2005;
Gray et al., 2004). This paper identifies a related
pollution hazard, namely that associated with the
presence of fine-grained (o10 mm) Hg-bearing particu-
late matter (HgPM10) dispersed across the former
mining and metallurgic areas.
2. Almaden geology and mining history
Almaden geology (Fig. 1) is dominated by a faulted
syncline in Palaeozoic siliciclastic shelf sediments in-
truded by mantle-derived mafic igneous sills, dykes and
diatreme-like breccias containing ultramafic xenoliths
(Hernandez et al., 1999). The main mercury-bearing
host rock is the Criadero Quartzite Formation which is
Silurian in age, up to 70m thick, and impregnated with
mercuric sulphide (cinnabar, HgS) and native mercury.
The igneous rocks are linked to transport of the mercury
from the mantle and can also be highly mineralised with
cinnabar and native metal (Higueras in Gibbons and
Moreno, 2002, and references therein). The ore grade in
these igneous rocks can be extremely high: in the Las
Cuevas mine (discussed below) liquid mercury could be
observed dripping from rocks impregnated with the
sulphide and native metal during underground visits.
Thus Almaden mercury ores are both high grade and
chemically simple so that residual contaminating mer-
cury phases in the mine dumps are mostly the sulphide
cinnabar plus some native mercury, with exposed rock
surfaces at both mine sites and processing plants
weathering to the yellow mercury sulphate schuetteite,
Hg3(SO4)O2 (Higueras et al., 2003).
Mercury mining is thought to have begun around
430BC, with geochemical data from Spanish peat bogs
recording the first significant amounts of anthropogenic
mercury (HgANT) at this time (Martinez-Cortizas et al.,
1999). These data show further notable increases of
ARTICLE IN PRESST. Moreno et al. / Atmospheric Environment 39 (2005) 6409–6419 6411
HgANT during Roman exploitation, especially once
Roman refining operations began, with the high grade
of the ore leading Pliny the Elder to comment that
Almaden produced the best vermilion used for dying
ceremonial robes. A post-Roman decline in mercury
emissions was reversed after the establishment of Islamic
rule, when metallurgy was practised: there are records of
1000 labourers being employed at Almaden in the 12th
century under the Arabic Empire (Hylander Lars and
Meili, 2003). Mercury was subsequently mined con-
tinuously at Almaden, which remained the world centre
of production and processing until towards the end of
the last century, when mining declined and ceased due to
environmental concerns over the metal’s toxicity.
Ninety per cent of total mercury production was at
Almaden itself, a town of around 7000 people with a
university mining school dating back to the 18th
century. The other mines lie east of Almaden, following
the narrow, faulted outcrop of the Criadero Quartzite
15 km southeast before curving back northwest around
the synclinal closure (Fig. 1). The two major sites of
Fig. 2. (a) Aludeles-type retorting furnace as used at Almadenejos du
from above the condensing chamber. (b) Unremediated waste dump o
furnace at Almadenejos. (d) Almadenejos guardhouse sample site, oc
recent mining activity here are the underground mine of
Las Cuevas (active in Roman times and from 1983 to
1999), and the opencast mine El Entredicho (main
extraction period 1983–1997) further southeast (Fig. 1).
Important ancient mines include the 17th–18th century
Vieja Concepcion and the slightly younger Nueva
Concepcion, which was active from 1798 to 1865 and
sited at Almadenejos, 11 km SE of Almaden (Fig. 1).
Both of these ancient mines served a mercury extraction
plant, also based at Almadenejos, where there were at
least 5 aludeles-type retorting furnaces used during the
17th–19th century before finally being decommissioned
in 1860 (Fig. 2a). Use of these furnaces involved simple
roasting of the HgS at temperatures above 300 1C,
dissociating the ore to form Hg vapour and SO2. The
mercury was then recovered by condensing the vapour
within enclosed pottery channels that were manually
cooled with slave labour and cold water. At the end of
the V-shaped pottery line was a chamber capped by a
metal lid where gas expansion induced further mercury
condensation. In more recent times, the processing was
ring the 17–19th centuries, with inset showing sample collection
utside the Las Cuevas underground mine. (c) Derelict retorting
cupied by Iberian pigs.
ARTICLE IN PRESST. Moreno et al. / Atmospheric Environment 39 (2005) 6409–64196412
done at the Almaden mine itself using propane ovens
which roasted the ore at 700 1C, with considerable
mercury losses to the atmosphere (Ferrara et al., 1998).
3. Sampling and analytical methodology
This study reports on mercury concentrations in
resuspendable airborne particulate matter in the Alma-
den mining region, with specific emphasis on the finest,
most easily inhaled particles. Samples were collected in
order to investigate a mine site, retorting plant, and the
background urban geochemistry of Almaden. Primary
mine site samples were collected from the unremediated
waste dumps lying outside the entrance to the aban-
doned Las Cuevas underground mine (Fig. 2b). Samples
from where the mercury ore was once processed were
collected from the derelict compound lying immediately
west of the village of Almadenejos (Fig. 2c): one sample
from a retorting site (top of the vapourisation chamber:
Fig. 2a), and another from 64m away in the dusty floor
of an old guardhouse at the gated entrance to the
compound (Fig. 2d). The latter was being used at the
time of sampling by domestic Iberian pigs (Sus scrofa)
sheltering from the extreme heat of the day (440 1C).
Samples were also taken of wild asparagus plants
(Asparagus acutifolius) growing within the Almadenejos
compound (Fig. 2c), as these are known to be collected
and consumed by the local human population (Higueras
et al., 2003, 2005). Finally, two samples of road dust
were collected from within Almaden town itself. The
first of these was taken from the University car park,
which lies immediately east (and commonly downwind)
from the Almaden mine and the processing plant. The
second site, a supermarket car park in Almaden town
centre, lies a further 700 m east.
Airborne resuspended dust was extracted from the
samples by placing each in a rotating enclosed drum
under a unidirectional air flow of 25 l min�1. The
rotating sample, disturbed by regular sliding and
cascading during each rotation, resuspends its finer
fraction into the airspace of the drum. The resuspended
aerosols buoyant enough to stay in resuspension thus
became drawn by the air current out of the drum, to
pass into a small chamber within which the largest and
heaviest particles were deposited. This initially redepos-
ited material was analysed for mercury and is referred to
as the coarse fraction. Those particles small and/or light
enough to be still capable of being carried by the air
current, continued their journey through the system,
entering a Negretti elutriation filter designed to allow
passage to only PM10 (particulate matter o10 mm) grade
material of average density. Those particles able to
travel through this barrier were finally collected on a
polycarbonate filter with 0.67mm porosity. Thus resus-
pended PM samples were separated into a coarse
fraction (deposited prior to entering the elutriation
filter), and fine PM10 (deposited on the polycarbonate
filter).
HgPM present within these samples were then
characterised chemically and morphologically at Cardiff
University using a combination of a low vacuum SEM
(JEOL5900LV) for single particle analysis via an energy
dispersive X-ray microanalysis system (EDX), ICP-MS
for whole sample analysis, and FESEM (XL30-FEG
SEM, Philips Electron Optics, NL) for imaging. Samples
were studied on SEM tubs with a sticky carbon cover
and then coated with gold/palladium (20 nm) prior to
examination via FESEM, or carbon in the case of the
SEM. Microscope conditions for backscatter imaging
were accelerating voltage of 20 kV, working distance of
10mm and with the gold foil aperture inserted.
Conditions for particle chemical analyses were working
distance 25mm, accelerating voltage of 20 kV, and the
beam spot size was 2. For whole sample chemical
analysis 500mg of each sample were digested using
concentrated nitric acid (Fisher Primar grade specific
gravity 1.48) before being analysed using a Thermo
Elemental X Series (X7) ICP-MS with a detection limit
for Hg equivalent to 18 ppb in the solid sample. The
instrumental precision was 5–6% for concentrations
o100 ppb (in solid), 1–3% for concentrations between
100 and 1000 ppb and 0.5–1% for concentrations
41000 ppb. Digestions were carried out in a CEM
MDS-200 microwave system, using CEM advanced
composite vessels with Teflon liners. The vessels were
pressure controlled, and pressure conditions, volume
and concentration of acid were varied until the optimum
conditions for complete dissolution of the samples were
found. The pressure increases to 80 psi over a period of
approximately 20 min producing a digestion tempera-
ture of approximately 180 1C. The digested samples were
then concentrated by evaporating the nitric acid, and
redissolving in 2 ml of 10% nitric acid. Samples were
diluted to a 20 ml volume using deionised (418MO)
water. One millilitre of each sample was combined with
a 50 ng g�1 thallium standard (1ml) and this solution
was made up to 10 ml with 2% nitric acid to be analysed
in the ICP-MS. Concentrations were quantified by
external calibration using synthetic mercury standard
solutions. Samples were analysed in order of expected
highest concentration and the ICP-MS glassware was
washed with 10% aqua regia for 300 s between samples.
Repeat analysis of blank solutions (blank for HNO3
36 ppb in the solid) and standards between the
unknowns indicated that memory effects due to
unremoved Hg occur. However these do not introduce
a contribution of more than 1.5% (relative) additional
Hg to the following sample and this additional
uncertainty does not affect the conclusions reached
below. It is more important to recognise that the
potential for some loss of Hg by volatilisation remains.
ARTICLE IN PRESST. Moreno et al. / Atmospheric Environment 39 (2005) 6409–6419 6413
Therefore the concentrations reported here are probably
under estimates of the true Hg concentration in each
sample but they can nonetheless reveal important trends
or differences between sample types—the main aim of
this study.
Finally, X-ray diffraction (XRD) analysis of the most
mercury-rich sample, i.e. that from above the Almade-
nejos retorting chamber were performed in a Bruker D-
5005 diffractometer with a CuKa radiation, a wave
length of l ¼ 1:5405 and a secondary graphite mono-
chromator. Analytical conditions were step size of 0.051,
with a 3 s timing per step, 40 kV and 30mA.
4. Results
Backscatter SEM study of the resuspended materials
from all sites reveals an obvious visual presence of
Fig. 3. SEM backscatter images of (bright) mercury-bearing particle
sample from Las Cuevas mine. Most HgPM are individual grains o
retorting site at Almadenejos; (b) ultrafine particles include spherical
sulphide particle probably derived from furnace; (d) euhedral HgPM
mercury-bearing particles in all samples, especially at the
Almadenejos processing plant and at the Las Cuevas
mine (Fig. 3a). Semi-quantitative SEM analyses of
individual HgPM10 from our samples indicate abundant
cinnabar (with, at Almadenejos, possibly its high
temperature polymorph metacinnabar) in all samples
(except sulphate encrustations) and the presence of
native Hg (Fig. 4). The latter is especially easy to find in
the Almadenejos sample taken from where the brick-
work and soils have become impregnated with metal
condensed from mercury-rich vapours rising through the
retorting plant. Several SEM analyses of HgPM10 at
Almadenejos record the presence of chlorine, and the
XRD results suggest, in addition to cinnabar, the presence
of the chlorine-phase eglestonite (Hg6Cl3O(OH)).
Although more sophisticated Hg speciation studies are
needed to confirm the phase chemistry, we are confident
from our data so far obtained that chlorine-bearing
s (HgPM) in resuspended dusts. (a) Low magnification view of
f cinnabar; (b–d) high magnification views of dust from above
condensed droplet of native mercury (centre); (c) cindery Hg-
phase grown in situ at the sampling site.
ARTICLE IN PRESS
Almadén Hg-bearing particles
0
2
4
6
8
10
12
14
0 10 20 30 40 50 60 70Hg (wt%)
S (
wt%
)
HgSnative Hg
Fig. 4. SEM analyses of sulphur and mercury in individual HgPM10, illustrating the presence of both sulphide and native metal.
T. Moreno et al. / Atmospheric Environment 39 (2005) 6409–64196414
HgPM10 are present in addition to the sulphur-bearing
mercury minerals and the native metal.
Figs. 3b–d illustrate the diversity of the mercury-
bearing particles, especially in the Almadenejos retorting
site dust which reveals a spectacular range of HgPM10
morphologies, from anhedral to euhedral, spherical to
acicular. There are small (2–8 mm) spherical metallic
mercury droplets condensed from the vapours, and
cindery, carious aggregates and crusts of mercury and
sulphur presumably derived from the furnace (Fig. 3b).
Polycrystalline, mercury- and sulphur-rich particles
comprising silica, clay minerals and a partial coating
of oxidised iron compounds are interpreted as soil and
brickwork fragments infiltrated by mercuric and sul-
phurous vapours. Elongate euhedral crystals of HgPM
have grown in situ (Figs. 3c–d), with other sulphur-
bearing Hg compounds instead forming thin condensate
films coating host minerals such as silica, clays and
calcite. The HgPM10 range down to nanometric specks
(Figs. 3b–d): the abundance of extremely fine-grained
material is particularly evident.
The Almadenejos guardhouse sample SEM back-
scatter imaging again reveals the floor dust to be
obviously rich in HgPM, although less abundant and
with less morphological range. All HgPM analyses show
the presence of sulphur +/� chlorine. HgPM can occur
as angular monomineralic grains but are normally found
embedded within clay mineral agglomerations in both
the fine (Fig. 5a) and coarse (Fig. 5d) dust fractions,
encased by thin (ca. 0.1 mm) clay coatings (Fig. 5b), or
attached to quartz or calcite (Fig. 5c). In the fine fraction
the mercury-bearing particles are usually much smaller
than 10mm in size, ranging down to the nanometric
specks seen in the previous sample.
In the Las Cuevas mine sample (Fig. 3a) HgPM
morphology is again less morphologically variable than
at the retorting site, with most particles being sulphides
and the remainder being native Hg, this being consistent
with the mixture of cinnabar and metallic ores mined
from here. The two urban car park samples show less
abundant, but still easily detectable, HgPM (sulphide
and native metal) in both coarse and fine fractions, with
monomineralic cinnabar grains, armoured multigrain
agglomerations, and clay-coated particles all being
present (Figs. 6a–d). As seen in the Almadenejos
guardhouse sample, clay armouring of the metallic
fragments is characteristic of these car park samples,
which have been physically reworked during resuspen-
sion and transport away from their original source. Such
armouring and clay coating presumably act to protect
the particles from physical attrition and chemical
attack during transport and gives them added durability
as atmospheric particulates capable of repeated
resuspension.
Fig. 7 presents the ICP-MS results for both coarse and
fine resuspended dust fractions, wild asparagus stem
material collected in the Almadenejos compound, and
from a pig hair collected from the Almadenejos guard-
house floor. The most immediate observation is the
extreme level of mercury contamination, especially
above the Almadenejos condensing chamber where
HgPM10 concentrations reach 94,000–95,000mg g�1.
The Las Cuevas mine sample shows much lower (but
still very high) Hg values which are close to 6000mg g�1
in the fine fraction. The other three sites show maximum
Hg PM10 values down to under 900mg g�1 for the
Almadenejos guardhouse, under 150mg g�1 for the
University car park, and around 50–100 mg g�1 for the
Supermarket car park sample. The second clear ob-
servation from these data is that there is always a
marked fractionation which favours concentration of
mercury in the finer fraction. In most of the samples
(Las Cuevas, guardhouse, both car parks) the coarser
resuspended fraction contains a little less than half
(43–48%) that of the PM10 fraction. In the Almadenejos
sample the fractionation is very much higher, this being
interpreted as due to the abundance of very fine
condensate within the Hg-impregnated soil (see discus-
sion). Fig. 7 also shows Hg levels within the asparagus
sample ranging from 35 to 65mg g�1, with Hg concen-
trations of 8–10mg g�1 in the pig hair collected with the
guardhouse sample.
ARTICLE IN PRESS
Fig. 5. HgPM in Almadenejos guardhouse dust sample under SEM. (a) Cinnabar grain partially coated with clay minerals;
(b) HgPM10 with fine clay coating; (c) cinnabar grain attached to calcite particle; (d) cinnabar grain embedded within a coarse
aggregate of quartz and clay minerals.
T. Moreno et al. / Atmospheric Environment 39 (2005) 6409–6419 6415
5. Discussion
Enhanced levels of mercury were to be expected in the
samples collected from the Las Cuevas mine and
Almadenejos processing plant. Levels of Hg contamina-
tion exceeding 1000 mg g�1 have been reported from
other mercury mines in Spain (Loredo et al., 1999;
Viladevall et al., 1999) and elsewhere in the world (e.g.
Kim et al., 2004), with the highest previously recorded
being 46,000 mg g�1 Hg in mine wastes downstream from
mines in SW Alaska (Bailey et al., 2002). Published
studies of anthrosols and calcines (retorted ore) speci-
fically from within the Almadenejos metallurgical
complex have reported mercury values of nearly 9000
and 34,000 mg g�1, respectively (Higueras et al., 2003;
Gray et al., 2004). Furthermore, as Kim et al. (2004)
have documented in a study of sieved samples from
Californian mines, mercury levels in mine wastes are
concentrated into finer size fractions. Thus, according to
data from the King, Knoxville and Sulphur Bank mines
in California, sieved samples in the size range 32–45 mm
can be broadly expected to contain mercury concentra-
tions around double those in the 125–250 mm range
(Kim et al., 2004).
Our data confirm widespread mercury contamination,
and demonstrate that the preference of mercury for finer
dust fractions continues down into particles fine enough
to be regularly resuspended by wind, with at least a
further doubling of mercury levels in the PM10 fraction
as compared to coarser aerosols. Thus, at Las Cuevas,
the PM10 sample contains nearly 6000mg g�1 whereas
coarser resuspended dust contains o3000mg g�1. The
Las Cuevas mine lies in an uninhabited area close to a
main road 8 km ENE from Almaden town and does not
appear to present an obvious threat to any nearby
population centre. However, ENE winds blowing
towards Almaden are not uncommon (Fig. 1), and
anyway the presence of such a major pollution hotspot
ARTICLE IN PRESS
Fig. 6. HgPM in University car park dust sample under SEM. HgPM usually occur aggregated with clays and quartz (a, b) but also as
individual grains (c) which usually are finely coated with clay (d).
T. Moreno et al. / Atmospheric Environment 39 (2005) 6409–64196416
in a semi-arid and windswept setting is undesirable,
particularly given the abundance of native mercury
present at this site during mining operations and its
proximity to a stream draining into the Valdeazogues
River (further discussed below).
The problem of toxic metal concentration in fine dust
is most extreme around former processing plants: at
Almadenejos mercury contents in the PM10 sample
increase nearly by an order of magnitude over those in
the coarser size fraction. Whereas the coarser fraction
yielded a ‘‘mere’’ 12,000–15,000mg g�1, the aerosols
collected on the polycarbonate filter comprised nearly
10% mercury, making this site to date the most
concentrated mercury pollution hotspot known on the
planet. Such extraordinarily high levels can be explained
by the fact that the processing buildings have been
pervaded by vapours emanating from the furnace over
many years, secondary Hg-minerals have condensed and
crystallised within the brickwork and soils, and possibly
that Hg has electrostatically bonded to certain clay
minerals that themselves are found in the inhalable size
range. Mercury concentrations of 19,500mg g�1 have
been reported from condenser soot from the New
Almaden mine in California (Kim et al., 2000, 2004),
and we would predict that metal levels in such soots
would be much higher in the PM10 fraction. Mercury
phases associated with these ore roasting systems
typically comprise a mixture of cinnabar and metacinna-
bar, elemental mercury, and mercury sulphates and
chlorides, and are known to be very fine grained
(Rytuba, 2003). This latter fact is well illustrated in
Figs. 4b–d which demonstrate how the PM10 sample
from above the Almadenejos condensing chamber, in
addition to many larger HgPM10, contains an abundance
of tiny mercury specks a few microns or less in size.
The data from the Almadenejos guardhouse indicate
that mercury levels diminish rapidly away from the
extreme levels at the retorting hotspot, although
concentrations of several thousandmg g�1 mercury
64m away obviously still means serious contamination.
ARTICLE IN PRESS
94769
5776
841
144101
9
50
13582
2687
407
6148
1
10
100
1000
10000
100000
Almadenejoscondensing
chamber
Las Cuevas Almadenejosguardhouse
University carpark
Supermarketcar park
Almadenejoswild asparagus
Almadenejospig hair
FINE
COARSE
pp
m
Average values
8-10Almadenejos guardhouse pig hair
35-65Almedenejos wild asparagus
43-53Almadén supermarket car park COARSE
101-101Almadén supermarket car park FINE
60-62Almadén University car park COARSE
136-152Almadén University car park FINE
350-464Almadenejos guardhouse COARSE
812-870Almadenejos guardhouse FINE
2 551-2 823Las Cuevas mine COARSE
5 676-5 875Las Cuevas mine FINE
12 050-15 113
94 137-95 401Almadenejos condensing chamber FINE
Almadenejos condensing chamber COARSE
µg/g Sample
Fig. 7. ICP-MS data from five sample sites, plus wild asparagus stem material growing in Almadenejos retorting compound, and pig
hair collected with Almadenejos guardhouse sample (two analyses per sample).
T. Moreno et al. / Atmospheric Environment 39 (2005) 6409–6419 6417
As at Las Cuevas, the inhalable PM10 fraction of the
guardhouse floor dust contains over double the mercury
of the coarser resuspended fraction. The enhanced
mercury levels present in the sample of pig hair from
the Guardhouse question the wisdom of rearing farm
animals in such a contaminated environment. Similarly,
the elevated levels of mercury within the wild asparagus
samples, an observation also reported in other studies
(Higueras et al., 2003, 2005), make them unsuitable for
consumption. The Provisional Tolerable Weekly Intake
(PTWI) for mercury set by the joint WHO/FAO Expert
Committee on Food Additives is 300mg person�1, of
which not more than 200mg should be methyl mercury.
Therefore, on that basis, the local inhabitants should eat
no more than 6 g of asparagus per person per week.
The two samples from the urban environment of
Almaden show HgPM levels much lower than at the
mine and retorting complex, although the University site
is the more contaminated of the two. The PM10 fraction
again contains over double the amount of mercury, and
so must also contain very many more numbers of
mercury-bearing particles than the PM410 mm fraction.
With levels exceeding 100mg g�1 in the fine fraction of
the Supermarket car park sample, even given the pilot
nature of this study it can reasonably be predicted that
at least low level Hg contamination is pervasive along-
side Almaden’s road system and probably elsewhere
throughout the town. HgPM10 measuring a few microns
across and protectively coated with low-density clays
will easily be repeatedly airlifted from one site to
another by wind transport and traffic resuspension.
Recent studies of surface waters in the Almaden area
have revealed extensive mercury contamination in
waters ponded in the abandoned opencast mine of El
Entredicho and the streams that drain the area (Berzas
Nevado et al., 2003; Gray et al., 2004). The environ-
mental hazard posed by fluvial runoff from exposed
mine and retorting wastes will be enhanced during dry
weather by toxic metal aerosols blown from pollution
hotspots and wind-transported to nearby streams and
ARTICLE IN PRESST. Moreno et al. / Atmospheric Environment 39 (2005) 6409–64196418
rivers. Inorganic mercury particles can be converted on-
site to extremely toxic organic compounds such as
methyl mercury (CH3Hg+): Gray et al. (2004) have
identified extensive bacterial methylation at the Alma-
denejos retorting site. Once deposited into the river
drainage system, further microbial methylation, which is
favoured by wet, anoxic, organic-rich environments, will
take place in aquatic conditions downstream.
The Las Cuevas mine and its waste dumps lie adjacent
to a stream with reported Hg levels of 41005 mg g�1
(Berzas Nevado et al., 2003), these concentrations
reducing to mostly below 100 mg g�1 downstream in
the Valdeazogues River, the main river draining the
area. The stream draining from the Almaden mine is
similarly highly contaminated, with up to 2300 mg g�1
Hg and up to 82 ng g�1 CH3Hg+ in the sediments, and
up to 13,000 ng l�1 Hg and 30 ng l�1 CH3Hg+ in the
water itself (Gray et al., 2004). Inevitably this methyl-
mercury, which is water soluble and highly prone to
biomagnification, is found in the food chain: the Berzas
Nevado et al. (2003) study of the Valdeazogues
ecosystem showed mercury in bivalves ranging from
2.8 to 4.1mg g�1, 20–40% of which being methylmer-
cury. This corresponds to a dry weight more than or
close to the maximum permissible level according to
WHO (0.5 mg g�1) for edible parts of the organism.
Alarmingly, mercury levels in crayfish eaten by the local
population have been recorded as exceeding 26 mg g�1
dry weight in some tissues (Higueras et al., 2005).
The Almaden area is unique in that it has been mined
for over 2400 years, and little effort has been made to
control or remediate pollution, so it is hardly surprising
that there remains a legacy of mercury contamination
that exceeds anything known elsewhere. Our study
indicates that much of the contaminating metal is
concentrated within PM10-size material. Contamination
of population centres close to old mines has been
reported elsewhere, such as in the Asturian town of
Mieres where urban dust samples containing mercury
contents 25 times higher than background occur 2 km
from the second largest mercury mine in Spain (Loredo
et al., 2003). PM10 dust from our least contaminated
(supermarket) site in Almaden town centre contains
mercury levels of 101 mg g�1, which is over 1600 times
higher than the average crustal abundance of the metal
(0.06 mg g�1: Wedepohl, 1995). In both Almaden
and Almadenejos the problem is likely to be com-
pounded by variations in mercury concentrations. Both
population centres lie downwind from mining activities,
in which context our data from the car park samples
support the reasonable expectation that background
values increase upwind towards the contamination
hotspots. In some places there are shafts that ventilated
underground mining operations (Ferrara et al., 1998),
which, added to the 30 m tall stack built for the ore
roasting ovens at Almaden, provide other HgPM
pollution hotspots and subsequent redistribution by
aeolian resuspension.
Mercury produces adverse physiological effects at
relatively low concentrations (Abrahams, 2002). The
main recognised problem at Almaden has been the
progressive poisoning of workers in direct contact with
mercury vapours. The combination of mercurial va-
pours and ultrafine HgPM10 at Almadenejos must have
produced an exceptionally poisonous working environ-
ment, especially with secondary mercury phases formed
in ore roasting systems being typically more soluble
(Rytuba, 2003) and therefore more bioavailable. When
mercury vapour is inhaled it is readily absorbed into the
bloodstream, and most enters the kidneys, from which it
is excreted as an Hg-protein compound (Salomons,
1995). The toxin has been estimated to have a half-life of
2 months, and if mercurial vapours are inhaled over long
periods chronic mercurial poisoning occurs, with devas-
tating effects on the central nervous system. Although
the effects of such poisoning are well documented (e.g.
Ishihara and Urushiyama, 1994), human responses to
long-term exposure of low mercury concentrations
produced by inhalation of HgPM10 remain unclear.
Like soil particles, which humans ingest at an average
rate of 10mg day�1 (Stanek et al., 1997), HgPM10 are
involuntarily brought into the body and many will be
fine enough to enter the deep lung and take part in
respiratory exchange. Our data indicate that local
residents in Almaden are exposed to inhalable dusts
containing mercury concentrations of over 100 mg g�1 in
their finest fraction. Because the adverse health effects
from such long-term exposure may be significant, future
epidemiological analyses and remediation programs
around former mercury mines should, in the light of
this study, pay especial attention to the character,
abundance, and distribution of HgPM10.
Acknowledgements
TM acknowledges the support of the Ramon y Cajal
scientific programme in the completion of this work. The
Director of the Escuela Universitaria Politecnica de
Almaden, Luis Mansilla, kindly provided logistical
support during sample collection. The authors thank
Minas de Almaden y Arrayanes S.A for the wind data
shown in Fig. 1 and original draft of Fig. 2a, and Josep
Elvira from CSIC for his help with the XRD analytical
results. The Cardiff ICP laboratory is supported by the
NERC, through Joint Infrastructure Fund award NER/
H/S/2000/00862.
References
Abrahams, P.W., 2002. Soils: their implications to human
health. The Science of the Total Environment 291, 1–32.
ARTICLE IN PRESST. Moreno et al. / Atmospheric Environment 39 (2005) 6409–6419 6419
Bailey, E.A., Gray, J.E., Theodorakos, P.M., 2002. Mercury in
vegetation and soils at abandoned mercury mines in
southwestern Alaska, USA. Geochemistry: Exploration,
Environment, Analysis 2, 275–286.
Berzas Nevado, J.J., Garcıa Bermejo, L.F., Rodrıguez Martın-
Doimeadios, R.C., 2003. Distribution of mercury in the
aquatic environment at Almaden, Spain. Environmental
Pollution 22, 261–271.
Ferrara, R., Maserti, B.E., Andersson, M., Edner, H.,
Ragnarson, P., Svanberg, S., Hernandez, A., 1998. Atmo-
spheric mercury concentrations and fluxes in the Almaden
distric (Spain). Atmospheric Environment 32, 3897–3904.
Gibbons, W., Moreno, T., 2002. The Geology of Spain. The
Geological Society of London, 649pp.
Gray, J.E., Hines, M.E., Higueras, P., Adatto, I., Lasorsa,
B.K., 2004. Mercury speciation and microbial transforma-
tions in mine wastes, stream sediments, and surface waters
at the Almaden Mining District, Spain. Environmental
Science and Technology 38, 4285–4292.
Hernandez, A., Jebrak, M., Higueras, P., Oyarzun, R., Morata,
D., Munha, J., 1999. The Almaden mercury mining district,
Spain. Mineralium Deposita 34, 539–548.
Higueras, P., Oyarzun, R., Biester, H., Lillo, J., Lorenzo, S.,
2003. A first insight into mercury distribution and speciation
in soils from the Almaden mining district, Spain. Journal of
Geochemical Exploration 80, 95–104.
Higueras, P., Oyarzun, R., Lillo, J., Sanches-Hernandez, J.C.,
Molina, J.A., Esbrı, J.M., Lorenzo, S., 2005. The Almaden
district (Spain): anatomy of one of the world’s largest Hg-
contaminated sites. The Science of Total Environment, in press.
Hylander Lars, D., Meili, M., 2003. 500 years of mercury
production: global annual inventory by region until 2000
and associated emissions. The Science of the Total
Environment 304, 13–27.
Ishihara, N., Urushiyama, K., 1994. Longitudinal study of
workers exposed to mercury vapour at low concentrations:
time course of inorganic and organic mercury concentra-
tions in urine, blood, and hair. Occupational and Environ-
mental Medicine 51, 660–662.
Kim, C.S., Brown Jr., G.E., Rytuba, J.J., 2000. Characteriza-
tion and speciation of mercury-bearing mine wastes using
X-ray absorption spectroscopy (XAS). The Science of the
Total Environment 261, 157–168.
Kim, C.S., Rytuba, J.J., Brown Jr., G.E., 2004. Geological and
anthropogenic factors influencing mercury speciation in
mine wastes: an EXAFS spectroscopy study. Applied
Geochemistry 19, 379–393.
Loredo, J., Ordonez, A., Gallego, J.R., Baldo, C., Garcıa-
Iglesias, J., 1999. Geochemical characterisation of mercury
mining spoil heaps in the area of Mieres (Asturias, northern
Spain). Journal of Geochemical Exploration 67, 377–390.
Loredo, J., Pereira, A., Ordonez, A., 2003. Untreated
abandoned mercury mining works in a scenic area of
Asturias (Spain). Environment International 29, 481–491.
Martinez-Cortizas, A., Pontevedra-Pombal, X., Garcia-Rodeja,
E., Novoa-Munoz, J.C., Shotyk, W., 1999. Mercury in a
Spanish peat bog: archive of climate change and atmo-
spheric metal deposition. Science 284, 939–942.
Rytuba, J.J., 2003. Mercury from mineral deposits and
potential environmental impact. Environmental Geology
43, 326–338.
Salomons, W., 1995. Environmental impact of metals derived
from mining activities: processes, predictions, prevention.
Journal of Geochemical Exploration 52, 5–23.
Saupe, F., 1990. Geology of the Almaden mercury deposit,
Province of Ciudad Real, Spain. Economic Geology 85,
482–510.
Stanek Jr., E.J., Calabrese, E.J., Barnes, R., Pekow, P., 1997.
Soil ingestion in adults—results of a second pilot study.
Ecotoxicology and Environmental Safety 36, 249–257.
Viladevall, M., Font, X., Navarro, A., 1999. Geochemical
mercury survey in the Azogue Valley (Betic Area, SE
Spain). Journal of Geochemical Exploration 66, 27–35.
Wedepohl, K., 1995. The composition of continental crust.
Geochimical and Cosmochimical Acta 59, 1217–1232.
Wilman, J.M., 2004. La amalgama y su influencia sobre el
cuerpo humano. Universidad de Valencia, Spain www.me-
dicina-naturista.net/salon_lectura/project5.pdf.